Production shows performance bugs not seen in staging and test. What change will make test and staging more representative and help prevent these bugs?

The correct answer is to increase the load on your test and staging environments. Performance bugs often manifest under specific load conditions that may not be replicated in typical test or staging setups. By simulating production-level traffic and usage patterns, you can uncover these issues before they impact live users. Deploying fewer or smaller changes to production are good practices for reducing risk but don't directly address the root cause of performance bugs appearing only in production. Deploying changes to a small subset of users (canary deployments) is a valuable strategy for mitigating the impact of production issues, but it's a post-deployment validation step rather than a preventative measure for finding performance bugs in pre-production environments.